AD841 or AD811

[Michael A]

Anybody know what an AD841 or an AD811 is? I'm trying to figure out some pre-amps that use these. They are 6 legged devices looking like a small transistor. The power supplied is positive only, 15V. I thought maybe they were dual transistors but my transistor checker refused to identify them.

Thanks,
Michael

 

[Michael A]

I'm pretty sure I'm reading these numbers right. They are pretty old devices with a datecode of 1980 if I read that right. Silver can with goldplated legs. I know the 841 gets pos 15V. The 811 I'm still trying to figure out. They are in little modules that slide into a rackmount unit so it is rather hard to trace them with power applied.

Michael

 

[bcarso]

I believe the 841 is a dual N-channel JFET, and certainly the AD811 you mention in a six-lead TO- case is a dual PNP. I used both of them around that date code time. AD ran out of part numbers, got out of the discrete business for the most part (except for the acquisition of PMI) and figured no one would care...

They were good parts too. I have a few pieces of the AD811 in a drawer.

 

[Michael A]

I did a search on dual N-channel JFET and came up with a LS841 at Linear Systems. They didn't have anything that was 811, but another module, I have 16 of them, has a MP311 in that position. LS has a 311 that is a dual NPN. What do you think? Are they the same things?

Thanks,
Michael

 

[bcarso]

Wait I think you might be right about the 811 being a dual NPN. The AD821 was the PNP now that my memory is jogged.

The AD840 at least was the dual FET I think, probably the 841 too. It was well-matched, medium voltage, not too much low-frequency noise, but otherwise unremarkable.

Although I don't think I have any of the FETs I might have some of the 811s and could determine the pinout and see if it corresponds to the Linear Integrated Systems parts.

EDIT: Yes LIS gives the LS840, 841, and 842 as subs for the AD parts, and the datasheet looks about as I remember, with things being characterized at 200uA.

Also LIS gives the LS311 as the sub for the AD811, and it is indeed an NPN dual: http://www.linearsystems.com/datasheets/LS310-3.pdf

 

[bcarso]

[quote author="mediatechnology"]OK, I'm pretty sure that I have an older Analog Devices data catalog that would have these. Let me look for them. Didn't realize that they re-used part numbers...[/quote]

Yes, I wonder whose decision that was. It does make things very confusing.

They were quite high on their discretes during that period (circa 1973 btw), with the typical our excreta don't stink attitude of AD of the period. Before I began to understand how FETs really worked I bought it too---there's an embarrassing gush I penned in the middle of an NSF proposal for a silicon target vidicon areal photometer where I extol the supposed virtues of the AD840 for a preamp. I also talk about the configuration of using half of it as an active drain load (less commonly seen than the source follower configuration, with half as the lower current sink) and make the very wrong assertion that it only has the voltage noise of a single device .

Well, we didn't get the money, not because of that blunder, but mostly because of political naïveté. But it was probably as well, since not too long after a sharp guy did a really good analysis, supported by experiment, of how truly lousy the vidicon was---that its readout noise was horribly worse than almost any preamp's*. By then a stopgap instrument was underway using a linear photodiode array. When that system was finally working guess what?---the intrinsic detector noise was again much larger than the preamp's. See a pattern? But I had fun, learned a lot, and some science was finally done.


*When I met the outspoken Andy Young at Table Mountain Observatory, I mentioned a paper of his about the vidicon photometry from one of the Mariner missions to Mars, in which he compares the quality unfavorably to some stellar photometry with gas-filled photodiodes in 1914. He laughed and said "Ahh that's the paper that got me fired from JPL!!"

One of the problems was that the innards of the vidicons got perturbed inelastically on takeoff, which screwed up the lab calibrations; another was spacecraft system noise from power supplies that hadn't been adequately suppressed; and then the science team needed to show the public pretty pictures, and not do recalibrations.

 

[Michael A]

Great story re: photonics.

So I've been trying to trace out this circuit a bit. I thought it might make a useful instrumentation amp if not audio. The differential input goes to the 841 configured as source followers with active loads from 2N5210 transistors. The output goes to the bases of the 811/311 part which also has some 2N5210s' around it. This might be a cascode though. I'm not very good at this. Before it gets to the 811 the signals are picked off to the non-inverting inputs of a unity gain configured LF353N. I haven't traced the output of that but I think it is ac coupled to the shields of the inputs. Above the 811 and 2N5210 is a MC1458 with one unit dc level shifted by the +12V supply to the non-inverting input and some network from the +15V supply going to the inverting. The other section seems to take the differential outputs from the collectors of the 2N5210s'. That's as far as I got. Seem possible? After the input area the signal goes to low pass and high pass filters and the gain setting.

Michael

 

[bcarso]

Not implausible from your description.